EMISSIONS DATA FROM TWO TUNNEL VENTILATED HIGH-RISE LAYER HOUSES IN NORTH CAROLINA Final Report for Site NC2B of the National Air Emissions Monitoring Study
Submitted to
John Thorne, Executive Director Agricultural Air Research Council C/O Crowell and Moring, LLP 1001 Pennsylvania Avenue, NW Washington, DC 20004
and
Bill Schrock, Environmental Engineer U.S. EPA Office of Air Quality Planning and Standards Mail Code E143-03, 4930 Page Road Durham, NC 27703
by
Albert J. Heber, Professor and NAEMS Science Advisor Agricultural and Biological Engineering, Purdue University 225 S. University St., West Lafayette, IN 47907 Phone: 765-494-1214 Email:
[email protected].
July 9, 2010
Contributors
Lingjuan Wang, Qianfeng Li Biological and Agricultural Engineering, North Carolina State University D. S. Weaver Labs,Campus Box 7625 Raleigh, NC 27695 Kaiying Wang School of Biosystem Engineering and Food Science Zhejiang University No. 268 Kaixuan Road Hangzhou 310029 The People's Republic of China Ilker Kilic, Albert J. Heber, Bill W. Bogan, Jiqin Ni, Lilong Chai, Claude Diehl, Juan Carlos Ramirez, Changhe Xiao, Agricultural and Biological Engineering Purdue University 225 S. University St. West Lafayette, IN 47907 Teng T. Lim Agricultural Systems Management University of Missouri 220 Agricultural Engineering Building Columbia, MO 65211
Acknowledgments This project was supported by the American Egg Board and the Agricultural Air Research Council. Citation Wang, K., I. Kilic, Q. Li, L. Wang, W.L. Bogan, J-Q. Ni, L. Chai, and A.J. Heber. 2010. National Air Emissions Monitoring Study: Emissions Data from Two Tunnel-Ventilated Layer Houses in North Carolina - Site NC2B. Final Report. Purdue University, West Lafayette, IN, June 18.
Table of Contents 1.
INTRODUCTION AND OBJECTIVES ............................................................................. 1
2.
CONFINED ANIMAL FEEDING OPERATION .............................................................. 1 2.1. 2.2. 2.3.
3.
Farm ................................................................................................................................. 1 Monitored Buildings ........................................................................................................ 2 Significant Events and Modifications .............................................................................. 4
MONITORING AND SAMPLING METHODS ................................................................ 4 3.1. General Approach ............................................................................................................ 4 3.2. Instrument Shelter ............................................................................................................ 4 3.3. Data Acquisition and Control System .............................................................................. 6 3.4. Monitoring and Recording Farm and Building Operation ............................................... 8 3.4.1. Animal Husbandry and Building Systems .............................................................. 8 3.4.2. Thermal Environmental .......................................................................................... 8 3.4.3. Building Airflow ..................................................................................................... 8 3.4.4. Biomaterials Sampling Methods and Schedule .................................................... 10 3.5. Particulate Matter Monitoring ........................................................................................ 10 3.6. Continuous Gas Sampling and Monitoring .................................................................... 10 3.7. VOC Sampling and Analysis ......................................................................................... 11 3.8. Documentation of Quality Assurance ............................................................................ 13 3.8.1. Oversight, Maintenance, and Calibration ............................................................. 13 3.8.2. Gas Sampling System ........................................................................................... 13 3.8.3. Gas Analyzers ....................................................................................................... 14 3.8.4. Particulate Matter Monitors .................................................................................. 16 3.9. Data Analysis ................................................................................................................. 17 3.9.1. Software ................................................................................................................ 17 3.9.2. Data Substitution, Validation, Correction and Uncertainty .................................. 18 3.9.3. Emission Calculations........................................................................................... 20
4.
RESULTS ............................................................................................................................. 21 4.1. Farm Production Information ......................................................................................... 21 4.2. Characteristics of Biomaterials ...................................................................................... 21 4.3. Environmental Conditions.............................................................................................. 21 4.3.1. Ambient Conditions .............................................................................................. 21 4.3.2. House Conditions .................................................................................................. 22 4.4. Ventilation Rate.............................................................................................................. 22 4.5. Particulate Matter Concentration and Emission ............................................................. 23 4.5.1. PM10 ...................................................................................................................... 23 4.5.2. PM2.5 .................................................................................................................... 24 4.5.3. TSP........................................................................................................................ 25 4.5.4. VOC Concentrations and Emissions..................................................................... 25 4.6. Hydrogen Sulfide Concentration and Emissions ........................................................... 26 4.7. Ammonia Concentration and Emissions ........................................................................ 28 4.8. Emission Data Completeness ......................................................................................... 28 4.9. Reconciliation with Data Quality Objectives ................................................................. 29
4.9.1. 4.9.2. 4.9.3.
Airflow .................................................................................................................. 29 Gas Emissions ....................................................................................................... 29 PM Emissions ....................................................................................................... 30
5.
SUMMARY .......................................................................................................................... 30
6.
REFERENCES .................................................................................................................... 30
7.
DEFINITIONS ..................................................................................................................... 31
APPENDIX A. MEASUREMENT VARIABLE LIST. ........................................................... 32 APPENDIX B: MAINTENANCE AND CALIBRATION ACTIVITIES. ............................ 37 APPENDIX C. GAS ANALYZER AND SPAN CHECKS ..................................................... 38 APPENDIX D: BIOMATERIALS CHARACTERISTICS .................................................... 40 APPENDIX E. DAILY MEAN DATA ...................................................................................... 41
1.
INTRODUCTION AND OBJECTIVES
The primary goals of the National Air Emissions Monitoring Study (NAEMS) were to: 1) quantify aerial pollutant emissions from dairy, pork, egg, and broiler production facilities, 2) provide reliable data for developing and validating emissions models for livestock and poultry production and for comparison with government regulatory thresholds, and 3) promote a national consensus on methods and procedures for measuring emissions from livestock operations. Emissions measurements were conducted at a total of 15 different house monitoring sites and ten open source sites in the continental US. The NAEMS was managed by Purdue University, in its role as Independent Research Contractor to the Agricultural Air Research Council. Purdue selected equipment and methods in consultation with the U.S. EPA, and subcontracted with other universities to operate the monitoring sites. North Carolina State University (NCSU) installed, maintained and calibrated equipment, collected samples, and conducted all other on-site activities. Purdue provided rapid feedback (generally within 2-4 business days) to catch aberrations in the data, and later conducted final processing of the data. Both NCSU and Purdue participated in reviews of the analyzed data. The overall objective of this report is to present the quality-assured measurements of ammonia (NH3), hydrogen sulfide (H2S), particulate matter (PM) and volatile organic compounds (VOCs) from two layer houses at the North Carolina egg layer facility. The specific objectives of the report are to: 1. Describe the farm, and the monitored buildings, 2. Describe the monitoring methods and quality assurance, and 3. Present tabulated daily averages of emissions. 2.
CONFINED ANIMAL FEEDING OPERATION
2.1.
Farm
This North Carolina egg layer facility (NC2B) consisted of six high rise houses with a total capacity of 618,000 hens, three natural ventilated houses, and an egg processing plant (Figure 1). There was one small swine farm and one small broiler production facility within 3.2 km of the site. Wastewater application fields were directly adjacent to the site, while the closest fields receiving actual solid manure from the farm were approximately 6.4 km away. The high rise houses were mechanically ventilated under the control of a computerized environmental control system. Birds were raised in low-density conditions per industry standards. A corn soy based diet was delivered to the birds in a trough at the front of the cages. Manure fell onto the curtain backed cages and then down into the first floor, where it was stored for up to one year.
1
Figure 1. Facility layout. Monitored buildings were houses 3 and 4. 2.2.
Monitored Buildings
Houses 3 and 4 (H3 and H4) were oriented north-south with 15.2 m spacing between them. (Figure 2). Each high-rise house was 175 m x 18 m, and contained 103,000 hens each in six rows of 4-tier A-frame cages in the upper floor. Each house had a sidewall height of 5.5 m and a first floor manure pit depth of 2.7 m. Hens were raised in low-density conditions (150 cm2/bird), and were molted according to standard industry practice. The producer weighed approximately 0.1% of the birds once every four weeks for an overall house average weight. During molting, the producer weighed the same birds on days 0, 4, 10, 12, 15, and 22. Hens were fed a corn/soy ration, with extra ingredients (crab meal, cookie meal, etc) added based on availability. Eggs were removed on conveyors. Egg production and water consumption were 2
automatically recorded daily, and feed consumption was checked daily via continuous monitoring using load cells under the feed bins. The second floor lights were shut off for 6-7 h each night. A standby generator was used to power critical systems during power outages. Each house was tunnel ventilated. Ventilation air entered the second floor through 36.5 m long air inlets that were centered on the east and west sides of the house. There were also eave inlets that admitted ventilation air during the winter. Finally, there were small openings that allowed air exchange from the outside into the pit. One of these openings was located on each sidewall, centered below the main inlets to the first floor. Each house had thirty-four, 122-cm diameter, 480-VAC, 3-phase, belt-driven exhaust fans (Choretime, Milford, IN) located on the east and west endwalls, with a spacing of 20 cm. Each endwall had eight fans in the second floor, and nine fans in the first floor. Each house had eight temperature sensors, and was ventilated in 11 stages. # Exhaust fan w/ current switch
# Exhaust fan w/ speed sensor (# = fan number)
House 3 8 11
5 9
4 9
3 10
2 1 11 11
15 14 6 4
13 12 1 5
11 7
10 3
7 6 10 10
17 16 8 2
9 8
House 4 25 11
Ventilation stages
East end of each house
24 23 10 10
34 33 8 2
22 9
32 31 6 4
21 20 19 18 9 10 11 11
30 29 28 1 5 7
27 26 3 8
West end of each house
Figure 2. Relay assignments to fans. Fan and relay numbers on outside and inside, respectively. Table 1. Fan numbers and stages. Bold type denotes fans activated at each stage. Stage Qty Fan ID 1 2 3 4 5 6 7 8 9
2 2+2=4 4+2=6 6+2=8 8+2=10 10+2=12 12+2=14 14+4=18 18+4=22
10
22+6=28
11
28+6=34
13, 30 13, 16, 30, 33 10, 13, 16, 27, 30, 33 10, 13, 14, 16, 27, 30, 31, 33 10, 12, 13, 14, 16, 27, 29, 30, 31, 33 10, 12, 13, 14, 15, 16, 27, 29, 30, 31, 32, 33 10, 11, 12, 13, 14, 15, 16, 27, 28, 29, 30, 31, 32, 33 9, 10, 11, 12, 13, 14, 15, 16, 17, 26, 27, 28, 29, 30, 31, 32, 33, 34 4, 5, 9, 10, 11, 12, 13, 14, 15, 16, 17, 21, 22, 26, 27, 28, 29, 30, 31, 32, 33, 34 3, 4, 5, 6, 7, 9, 10, 11, 12, 13, 14, 15, 16, 17, 20, 21, 22, 23, 24, 26, 27, 28, 29, 30, 31, 32, 33, 34 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34
Manure fell onto the curtain backed cages and then down into the first floor, where it was stored for up to one year and dried with 21 recirculation fans in the pit. It took several months to load
3
out of 750,000 kg of manure through the load-out doors located on the sidewalls between houses 3 and 4. The front door was located about 45.7 m from the front of the house. 2.3.
Significant Events and Modifications
The animal management and genetics for houses 3 and 4 remained the same during the study. Several 2-5 min long farm power outages occurred on 10/18/07. All ceiling-mounted manure drying fans were connected to fan stage 4 on 3/26/08. The fan staging in H3 was modified on 4/9/08. Fan belt replacement activities for all houses began on 5/21/09. The manure was cleaned out from each house three times during the monitoring period. House 3 was cleaned out March 9-14, 2008, March 21-26, 2009 and July 21-27, 2009. House 4 was cleaned out March 8-13, 2008 and March 19-24, 2009 and July 21-27, 2009. These activities caused abnormally high inlet PM concentration measurements because of the proximity of the Beta Gauge monitor to the manure loadout activity. 3.
MONITORING AND SAMPLING METHODS
3.1.
General Approach
Equipment installation and preliminary testing began on 5/21/07 and was completed on 9/25/07. The site setup and equipment installation followed approved site monitoring plan, a quality assurance project plan, and instrument or method-specific standard operating procedures. The monitoring period began on 9/25/07 and concluded on 9/30/09. Target pollutants for this site were NH3, H2S, PM (PM10, TSP, and PM2.5), and VOC. Appendix A lists the target pollutants, and all measured supporting variables and metadata monitored at the site. The monitoring schemes for the two structures are shown in Figures 3 and 4Error! Reference source not ound.. Table 2 lists the major equipment used at site NC2B, including the model, manufacturer and instrument specifications. A total of 220 and 116 visits were made by NCSU personnel during years 1 and 2 of the monitoring period. Remote checking via the internet was conducted by the NCSU and/or Purdue on a near-daily basis. The Science Advisor audited the site on 9/11/07 and 1/5/09. The Environmental Protection Agency (EPA) conducted site audits on 10/28/08 and 9/24/09. 3.2.
Instrument Shelter
The on-farm instrument shelter (OFIS) was located between houses 3 and 4 towards the east ends of the houses. Heated raceways were used to connect the OFIS with each house, to avoid condensation in the sampling tubing where it was exposed to inlet temperatures. The raceway temperatures were monitored continuously.
4
5
Table 2. Major instrumentation. Analyzer/Instrument INNOVA 1412 Multi-gas analyzer TEI 450i H2S analyzer Environics 4040 dilutor TEOM 1 (house 1) TEOM 2 (house 2) TEI FH 62C14 (Beta Gauge)
Serial number 710-197 709220676 2922 26512 26370 E-1275
The OFIS was supplied with 3-wire, single-phase, 120/240-volt, 100 A at 240 V power by the farm and connected to the external pullout switch at the OFIS. A copper ground rod was installed at the location of the OFIS and connected to the OFIS ground. The HVAC system of the OFIS maintained inside temperatures within the operating range of the analyzers, and created a positive pressure with a filtered outside air intake to minimize entry of unfiltered outside air. The temperature and differential static pressure in the OFIS were monitored with a thermocouple near the instrument rack and a pressure sensor. One set of gas analyzers (Table 2) in the OFIS measured gas concentrations as the gas sampling system (GSS) sequenced through all the gas sampling locations (GSLs). Vacuum pumps and controllers for the PM monitors were located in the OFIS. A personal computer collected all site monitoring data using a data acquisition and control program AirDAC. 3.3.
Data Acquisition and Control System
The data acquisition and control system consisted of a personal computer, custom software (AirDAC) written in a commercial programming language (LabVIEW, National Instruments, Austin, TX), distributed I/O hardware (National Instruments FieldPoint modules), and Universal Serial Bus (USB) devices by National Instrument (NI) and Measurement Computing (MC,, Norton, MA). The NI FieldPoint (FP) modules and MC USB devices (Table 3) were selected and configured to acquire data for all the on-line measurement variables (Appendix A). The 16-channel NI FP-DO-401 digital output module was used to control: 1) sequential switching of multiple gas sampling lines, 2) the raceway heating system, and 3) the GSS cooling fan. Serial communication (RS232) was used to acquire data from the multi-gas monitor and calibration variables (calibration time, gas concentration, etc.) from the gas dilutor. Voltage or current analog signals from various analyzers and sensors were connected to FP-AI-112 modules. Type T thermocouples were connected to FP-TC-120 modules. Digital signals from current switches and relays were connected to the MC USB DIO96H device. Voltage pulses from proximity sensors used to measure fan rotational speed were detected by the MC USB 4303 Counter. AirDAC averaged the signals (after conversion to engineering units) over 15-s and 60-s intervals and recorded the means into two separate computer files. All real-time data were displayed in tabular and graphic forms for on-site or remote (pcAnywhere, Symantec, Mountain View, CA) viewing (Ni et al., 2009; Ni and Heber, 2010). Measurement alarms, data collection notifications, data files, graphs and statistics of the daily data sets, and modified configuration and fieldnote files were automatically emailed to several recipients after midnight. 6
Wind sensor RH/T sensors
House 3 (18 x 175 m)
GSL
Thermocouple
P port
Activity sensor PM monitor OFIS Heated raceway Loadout door
House 4 (18x 175 m) Air inlets Exhaust fans
Walkway
N Open anemometer S Solar sensor
Legend:
Figure 3. Overhead view of sensor and air sampling locations at the monitoring site. Wind sensor
S
RH/T sensor
N
House 4 Heated raceway
House 3
Weather tower Activity sensor
Air plenum Air inlet
OFIS Open anemometer Legend:
P port
Thermocouple
Exhaust fan
PM monitor
GSL S Solar sensor
Figure 4. End view of monitoring plan for continuous emission testing at houses 3 and 4.
7
S
Table 3. Data acquisition hardware configuration for NC2B. Manufacturer and model I/O type # units # channels/unit Notes NI FP-AI-112 Analog input 4 16 Single-ended, 16-bit NI FP-TC-120 Thermocouple 2 8 NI FP-DO-401 Digital output 1 16 2 A at 10-30 VDC MC USB 4303 counter Count input 5 10 MC USB DIO 96H Digital input 1 96 3.4.
Monitoring and Recording Farm and Building Operation
3.4.1. Animal Husbandry and Building Systems Infrared motion sensors (activity sensors) were located to monitor movement of birds and workers in the house, with a total of four such sensors positioned in each house. An activity sensor was used to monitor researcher activity in the OFIS. Weekly layer inventories, egg production, feed and water consumption, egg production and characteristics and bird mass data were collected from the farm’s computer system. Layer inventories and weight were verified quarterly by the the site personnel. 3.4.2. Thermal Environmental Weather data was collected using a solar radiation shielded capacitance-type relative humidity and temperature probe (RH/T) (Model RHT-WM, Novus Automation, Porto Alegre, Brazil), a pyranometer (Model LI-200SL, LI-COR, Lincoln, NE) and a cup anemometer (Wind Sentry, RM Young, Traverse City, MI), which were attached to a 10-m high tower placed between houses 2 and 3. For the house environment conditions, capacitance-type RH/T probes were located at the PREFs (fan 13 in each house), and in one cage of each house. The cage probe was placed in the east half of the house, approximately centered, in an empty. Thermocouples (TC) were used to measure temperatures at four locations in the center of the house, one gas-sampling point (fan 4) in each house, the heated raceways, the OFIS itself, and the instrument rack. 3.4.3. Building Airflow Fan rotational speed and operation was monitored using a magnetic Hall-effect sensor (speed sensor) or current switches to monitor the on/off status of a fan based on its current draw. The speed sensors were mounted to detect the rotational speed in revolutions per minute (rpm) of either the fan shaft or the fan pulley. The digital signal from the speed sensor was converted into a frequency measurement with a counter module in the data acquisition system. Each fan already had a speed sensor that provided a 0-5 VDC signal to the farm’s environmental control computer; however, these sensors could not be interfaced to the DAC system in the OFIS. Therefore, a second speed sensor (Cherry Sensors, Model MP100701) was installed on fans 13 and 30 in each house, and on two fans of each higher fan stage (stages 2-11). Current switches monitored the on/off status of the other fans. 8
Static pressure differences were measured across the west and east endwalls of each house with differential static pressure sensors (Model 260, Setra Systems, Boxborough, MA). Static pressure differences were also measured across the walls through which the egg conveyor traveled between houses 2 and 3, 3 and 4 and 4 and 5.The outside port was located against the outside wall near the ventilation fans of the east wall. Static pressure in the OFIS was measured with the same type of sensor, to ensure that positive pressure was maintained. Impeller anemometers (Model 27106RS, RM Young, Traverse City, MI) were installed on the outlet of each of the PREFs (one per house). In-situ airflow measurements were conducted with a 122-cm field-portable fan tester (Fan Assessment Numeration System or FANS, University of Kentucky, Lexington, KY), which was described by Gates et al. (2004). The field data was used to develop equations that would calculate airflow as a function of differential pressure and fan rotational speed, and to assess the uncertainty in airflow predictions. A total of 196 in-situ fan tests with three replications were conducted during November 2007, June and July 2008 and July 2009. Each fan was tested at least once during the four testing periods. In November 2007, 100% of the fans in houses 3 and 4 were tested. In June 2008, 65% of the fans were tested, and in July 2009, 97% fans were tested. In July 2008, three selected fans were tested to measure air flow rates under three static pressure settings to determine the fieldmeasured fan curves as compared with published fan curves. The airflow curves of the all fans (Model 38264-4822, Chore-Time, Milford, IN) were obtained from the Bioenvironmental and Structural Systems (BESS) Lab at the University of Illinois at Urbana-Champaign (BESS, 2003; BESS, 2004). Each performance record consisted of airflow (Q1) measured at several static pressures (P1), and at a relatively constant speed (N1 = 600). The BESS fan curve was adjusted to the mean speed of the fan tests (N2), which was 570 rpm. The new, speed-indexed baseline curves were derived using the first (Q2 = Q1(N2/N1)) and second (ΔP2 = ΔP1(N2/N1)0.5) fan laws, where Q2 is the speed-adjusted BESS fan curve at speed N2. The speed-corrected airflow prediction model is Q4 = (aΔP4 + b)·(N4/N2)·Q2, where ΔP4 and N4 are measured fan static pressure and speed. For a given test using the portable tester, the model is Q4 = (a·ΔP3 + b)·(N3/N2)·Q2, where ΔP3 and N3 are the measured fan static pressure and speed during the fan test, and the fan degradation factor k = a·ΔP3 + b. The values for the coefficients a and b were those which minimized the sum of square differences between Q4 and Q3 for all the valid fan tests within a speed regime. The resulting fan model is shown in Table 4. Fans were assigned to a sampling stream based on their proximity to the three sampling locations in each house. For each house, fans 1 through 8 constituted stream 1, fans 18-34 made up stream 2, and stream 3 was fans 9 to 17. The airflow for each stream was calculated by summing the individual airflows for all fans in the stream.
9
3.4.4. Biomaterials Sampling Methods and Schedule All analyses of biomaterials were performed by an independent laboratory (Midwest Laboratories, Omaha, NE).Water was evaluated based on total N analysis of two samples collected on 1/16/08. Table 4. Fan airflow model. Reference Polynomial coefficients of Q2=f(ΔP2) at speed N2 a3 a2 a1 a0 speed (N2) 570 1.38E-05 4.86E-04 6.09E-02 1.10E+01
Coefficients of k b1 b0 3.94E-03 0.825
Manure surfaces were sampled from 1/16/08 to 8/13/09 (Table F1) to determine pH, moisture content and total ammoniacal nitrogen. There were 6 manure “windrows” in each house on the first floor. Each windrow was 166-m long. A block random sampling procedure was used to take the manure surface samples. Each windrow was divided into 68, 2.44-m long sections for a total of 408 sections per house. The computer program randomly selected 40 numbers representing 40 sections to be sampled. Forty samples of approximately equal weight was randomly collected from each section. The 40 samples were mixed thoroughly and 12 to 15 samples (about ½ kg each) were taken from the mixture and sent to the lab for analysis. A total of 188 samples were analyzed. Loadout manure was sampled during each full cleanout of the houses and were analyzed for pH, moisture content, total N, and ammoniacal N. During cleanout, 12 random samples per house were taken from either the truck or the skid loader. The volume of manure removed from each house was documented by the producer. 3.5.
Particulate Matter Monitoring
Real-time PM monitors (TEOM Model 1400a, Thermo Fisher Scientific, Waltham, MA) were located immediately upstream of fan 13 to continuously measure exhaust PM (Figure 4). Fan 13 in each house was referred to as the primary representative exhaust fan (PREF). A beta attenuation PM monitor (Beta Gauge Model FH62C-14, Thermo Fisher Scientific, Franklin, MA) continuously measured house inlet PM concentration. The Beta Gauge was enclosed in a protective outdoor enclosure and located at the inlet gas-sampling location near the SW corner of house 3 (Figure 4), near the ventilation air inlets. At any one time, the sampled PM size class was either PM10, PM2.5 or TSP at both TEOMs and the Beta Gauge. The PM2.5 size class was measured in January-February, 2008, October, 2008 and July-August, 2009 for 6 to 19 d each time (Table 5). The TSP inlet heads were placed on the TEOMs and Beta-Gauge for nine, 5 to 16 d periods. The PM10 concentration was measured at all other times. 3.6.
Continuous Gas Sampling and Monitoring
Air samples for continuous gas measurements were collected from multiple gas sampling probes with a custom-designed GSS. Each probe was connected to the GSS with Teflon tubing. Tubular raceways between the OFIS and the monitored buildings protected the sampling lines and data 10
signal cables. The sampling lines were wrapped with insulation and heated inside the raceways and at other locations vulnerable to cold air to prevent condensation inside the tubes. Three gas sampling probes were placed in each house, 0.5 m in front of the exhaust fans at a height equal to the fan hubs (Figure 4). Gas sampling probes A and B were located in front of the inlets of stage 1 fans 13 and 30 the east and west ends of the houses on the first floor. Sampling probe C was located in front of stage 9 fan 4 on the east end of the house on the second floor (Figure 4). The inlet air was sampled near the house 3’s SW corner, 2 m from the south wall and 2 m from the west end (Figure 4). Each exhaust location was sampled individually for 10 min. The ventilation inlet location was monitored at least twice daily, originally with a 20-min sampling period. In January, 2008, gas concentration data at each sampling location was studied to determine whether equilibrium occurred within the sampling periods. A statistical analysis confirmed that 10 min was sufficient for the exhaust GSLs, but that 30 min was required for the house inlet. The inlet sampling period was therefore increased from 20 to 30 min on 1/28/08. One set of gas analyzers in the OFIS was used to sequence through all the GSLs. Hydrogen sulfide was measured with a fluorescence H2S analyzer (TE Model 450C, Thermo Fisher Scientific, Waltham, MA). Concentrations of NH3 and CO2 were measured with a photoacoustic infrared multi-gas monitor (INNOVA Model 1412, LumaSense Technologies A/S, Ballerup, Denmark). 3.7.
VOC Sampling and Analysis
Grab samples of VOC were collected at fan 13 on the first floor and fan 4 on the second floor in house 4 (Table 3), using methodology based on methods TO-15 and TO-16. Sampling was conducted with 6-L stainless-steel canisters (TO-Can, Restek Corp, Bellefonte, PA), equipped with ¼″ bellows valves (Swagelok SS4H) and 207-kPa vacuum gauges. Sampling trains contained flow controllers (Veriflo Model 423XL, Parker-Hannifin Corp., Richmond, CA) with 2- to 4-sccm critical orifices and 7-µm in-line stainless steel filters. Flow controllers were pre-set to a constant flow rate of 3.4 mL/min. Canister sampling was conducted for 24 h, and canister pressures were recorded at the beginning and end of the sampling periods for the calculation of total sample volumes. Sampling was conducted seven times between 4/12/09 and 9/18/09, with duplicate samples typically collected at each location. All canisters were cleaned and passed QC before sample collection. Canister samples were analyzed at Purdue University’s Trace Contaminant Laboratory. The canisters were pressurized to +207 kPa with ultrapure N2, and transferred to TDS tubes (Carbotrap 300, Supelco, Bellefonte, PA). The pressurized canisters initially yielded sample flows of 50 mL min-1 during sample transfer to tubes. Canister heating was introduced when a canister pressure decreased to 13.8 kPa to ensure maximal transfer of nonvolatile components. Samples were analyzed on a thermodesorption-gas chromatograph-mass spectrometer (TDS-GCMS), consisting of a gas chromatograph (Model 6890, Agilent Technologies, Palo Alto, CA) coupled with a Model 5795 mass spectrometer detector (Agilent Model 5795) and equipped with a thermal desorption system (Model TDS-G, Gerstel, Baltimore, MD) and a cooled injection system (Gerstel CIS). The GC-MS passed a leak check prior to analyzing each set of samples. Compounds were separated on a 60 m x 0.25 mm x 1µm column. The detector utilized the full 11
Table 5. Sampling schedule for PM10, TSP and PM2.5. Time and day, hr:min-m/d/y Start Stop
9/24/2007 1/16/08 2/4/08 3/26/08 4/4/08 5/12/08 5/28/08 8/7/08 8/21/08 8/21/08 9/11/08 10/17/08 10/23/08 10/24/08 10/30/08 1/9/09 1/15/09 2/26/09 2/27/09 3/4/09 4/2/09 4/10/09 6/4/09 6/11/09 7/24/09 8/6/09 8/6/09 8/7/09 8/17/09 8/20/09 8/27/09 9/15/09 9/15/09 9/22/09
PM10
1/16/08 2/4/08 3/26008 4/4/08 5/12/08 5/28/08 8/7/08 8/21/08 9/11/08 9/11/08 10/17/08 10/23/08 10/24/08 10/30/08 1/9/09 1/15/09 2/26/09 2/27/09 3/4/09 4/2/09 4/10/09 6/4/09 6/11/09 7/24/09 8/6/09 8/17/09 8/7/09 8/17/09 8/20/09 8/27/09 9/15/09 10/6/09 9/22/09 10/6/09
Test duration, d TSP
PM2.5
114.5 18.9 51.1 9.1 37.9 16.0 71.1 14.0 20.9** 20.9* 36.0 5.8 0.9 6.0 71.2 5.8 42.0 1§ 4.9§ 28.9§§ 8.0 55.1 7.0 43.0 12.8 10.9**
1.0† 10.2* 2.9 6.8 19.2 20.9* 6.8** **
14.2 628
Totals *All except inlet **Only inlet †Only H4 upstairs ‡All except H3 §H4 TEOM collocated with H3 §§H3 TEOM relocated to H4 upstairs
12
98
49
scan mode covering masses from 27-270 Daltons in 8 scans/s. The MS quad hold temperature was 150ºC, and the MS source hold temperature was 230ºC. The analytical results were analyzed by ChemStation, and all integrations were manually checked. This method used an external standard compound for instrument monitoring and QA to avoid losses of low-molecular-weight analytes that would occur when purging solvent used with internal standard(s). All TDS tubes were cleaned with a tube conditioning system (Gerstel TC-2 TDS) for 3.5 h at 350ºC prior to each use. Table 6. Analyte sampling locations. Analyte House Sampling location* 3, 4 GSL-A: fan 13, east end NH3 3, 4 GSL-B: fan 30, west end H2 S 3, 4 GSL-C: fan 4, east end CO2 3, 4 INLET: 2 m from south wall, and 2 m from west end of air inlet of 3, 4 Fan 13,3 east end, underneath egg conveyor belt PM2.5 house 3, 4 INLET: Beta Gauge 2 m from south wall, and 2 m from west end PM10 4 Same as GSL A (upstairs) VOC of air inlet of house 3 TSP 4 Same as GSL C (downstairs) *Gas sampling probes were located at fan hub height, suspended from the ceiling.
Qty 2 2 2 1 1 1 2 2
Response curves were generated at both the beginning and the end of the VOC analysis period. The response curves of all chemical standards reach good linearity as 55% of the response curves had R2 > 99% and over 98% had R2 > 95%. Toluene was used as an external standard that was analyzed during each batch of samples to assure quality. The relative bias and standard deviation of 97 toluene checks were -4.3% and 18.8%, respectively. The uncertainty of the mean of duplicate field samples was calculated as 27%, based on the toluene checks. 3.8.
Documentation of Quality Assurance
3.8.1. Oversight, Maintenance, and Calibration North Carolina State University personnel visited the site frequently during the first few months of the study; that frequency declined as site operation became more routine. A total of 220 and 116 visits were made during years 1 and 2 of the monitoring period. The NAEMS Science Advisor audited the site on 9/11/07 and 1/5/09. The Environmental Protection Agency (EPA) conducted site audits on 10/28/08 and 10/24/09. Various site maintenance and calibration activities were conducted by site personnel (Appendix B). Specific quality assurance tests of the GSS, gas analyzers and other sensors are discussed below. 3.8.2. Gas Sampling System Two types of GSS leak tests were conducted. The first test examined GSS integrity, by briefly creating a “dead head” against the pump by closing all solenoid valves, while measuring exhaust airflow with a portable rotameter, and recording the leakage flow with the GSS mass flow meter. The second test consisted of monitoring GSS flow and pressure after manually setting AirDAC to sample from a particular GSL and plugging the GSL’s gas sampling probe, which created a 13
GSS manifold vacuum of about -75,000 Pa or 0.26 atm. Preliminary tests indicated that GSS flows under dead-head conditions that were 10% or less (
